RESUMEN
Early detection of some cancers (especially of the colon) may be achieved by nuclear magnetic resonance (NMR) spectroscopy applied to whole blood samples. Analysis by conventional Fourier signal processing techniques of the data so obtained has proved to be unreliable because of noise within the associated spectrum. This paper describes a neural network approach to analysis of the NMR data. At the present time, this method has proved to be highly reliable in differentiating between patients with an without colon cancer.
Asunto(s)
Neoplasias del Colon/diagnóstico , Espectroscopía de Resonancia Magnética , Redes Neurales de la Computación , Neoplasias del Colon/sangre , HumanosAsunto(s)
Antineoplásicos Fitogénicos/aislamiento & purificación , Ácido Gálico/análogos & derivados , Plantas Medicinales/análisis , Animales , Antineoplásicos Fitogénicos/uso terapéutico , Supervivencia Celular/efectos de los fármacos , Ácido Gálico/aislamiento & purificación , Ácido Gálico/uso terapéutico , Leucemia Experimental/patología , Ratones , Fitoterapia , Inhibidores de la Transcriptasa InversaRESUMEN
An aqueous alcoholic extract of fresh flowers of Yucca glauca Nutt. showed striking antitumor activity against B16 melanoma in mice. Systematic fractionation of the extract by means of solvent extraction and gel permeation chromatography led to separation of two galactose containing polysaccharide fractions with marked inhibitory activity against B16 melanoma in mice. The extraction, fractionation, purification and preliminary characterization of the active polysaccharide fractions are described. The materials showed no activity against L1210 or P388 leukemias in mice. The implications of the findings for searches for both natural and synthetic anticancer agents are discussed.